About Me

I’m a third-year undergraduate student majoring in Mathematics and Physics at Tsinghua University. My research interests lies in ML/CV, including generative models, VLMs, Video Gen/World Model and training dynamics, ICL. I prefer the principles for understanding neural networks and building next-generation AI.

Now I’m looking forward to 27fall phd positions.

Publications

(*Equal Contribution, #Corresponding Author)

dise Astra: General Interactive World Model with Autoregressive Denoising
Yixuan Zhu*, Jiaqi Feng*, Wenzhao Zheng, Yuan Gao, Xin Tao, Pengfei Wan, Jie Zhou, Jiwen Lu
The Fourteenth International Conference on Learning Representations (ICLR), 2026
[Paper] [Code] [Project Page]

We introduce Astra, an interactive world model that delivers realistic long-horizon video rollouts under a wide range of scenarios and action inputs.

dise One-Forcing: Towards Stable One-Step Autoregressive Video Generation
Jiaqi Feng, Justin Cui, Yuanhao Ban, Cho-Jui Hsieh
arXiv preprint arXiv:2605.23458, 2026
[Paper] [Code] [Project Page]

One-Forcing enables stable one-step autoregressive video generation by augmenting DMD-based causal distillation with a shared noised-latent adversarial critic.